Abstract
BACKGROUND: This study investigated the accuracy of the "PLT Clumps?" flag triggered by different channels on the Sysmex XN haematology analyser in identifying two sample states and to optimize the platelet review strategy for cancer patients. METHODS: 570 samples flagged "PLT Clumps?" in CBC + DIFF mode were analysed. After excluding clots, retesting was conducted in CBC + DIFF + PLT-F mode. The accuracy of the "PLT Clumps?" flag and its correlation with PLT aggregation (PA) or fibrin precipitation (FP) were evaluated. Low-platelet count samples were collected for a second blood draw to confirm whether pseudothrombocytopenia (PTCP) was present. The incidence and positive predictive value (PPV) of the "PLT Clumps?" flag across different tumours were also analysed. RESULTS: Among 85 verified cases (PA = 17, FP = 63), the overall PPV of the "PLT Clumps?" flag was 14.9% (85/570) in CBC + DIFF mode, 3.5% in WNR channel only, 73.9% in WDF channel only, and 94.4% in dual-channel. In CBC + DIFF + PLT-F mode, the overall PPV and NPV were 98.5% (65/66) and 97.0% (484/499), respectively. PA and FP accounted for 73.9% (17/23) and 26.1% (6/23) of the PLT-F-only flagged samples, respectively, and all WDF-only and dual-flagged cases were triggered by FP. The incidence of PTCP in samples flagged for "PLT Clumps?" in CBC + DIFF + PLT-F mode was 80%. Samples from patients with hepatobiliary tumours showed highest flagging rate (20.4%) but a lower PPV. CONCLUSION: The correlation between "PLT Clumps?" flags originating from different channels and two common preanalytical factors interfering in PLT counting was analysed, and the platelet review strategy for cancer patients was optimized.